error manifold

[ipotemusa]

Good morning to all!!! thanks to giuseppe of the noble for the beautiful words of opening of the thread.
the most beautiful part of the event was definitely that happened during the break.
"the matrix introduce ceschi1959 and www.cad-3d.net":finger:
But we managed to do what we had predicted. that is to please those who did not know nx but also those who knew him already. and we were not superficial.
now for lack of time I will delay to publish slides and results because of an upcoming fair in the states but I promise to publish them.. ..maybe even with some videos to support those who could not come.
as the matrix rightly said the matrix systems based on unix (linux and mac) have down the same performance. (we chose as a comparative the management of a set of 7gb) the only rock that pay these systems call them "alternative" are those linked more to the graphic aspect, i.e. the radial pop up and transparencies and the file opening mask "of old style motif".
for the rest nx is always him!!!! I mean that once inside the graphical interface is always the same that we know well and therefore the user finds himself at home. in fact it turns with x11, but as we have had way to explain and that I advance here quickly the x11 is not an emulator!:bekle:
It is a user interface that deals with the system's drive with the user and that is, it deals with the management of the mouse and keyboard. open motif libraries from the name itself are bookcases that instead deal with the graphic aspect of when nx interfaces with the system (file --> open etc etc). so nx on mac does not turn with an emulator but is installed in the "dark side of the moon" i.e. exploits the unix soul of mac. we also found an official siemens document in which it is said that this solution is under examination to see how the market responds. If there was a good answer, it would be rewritten for Mac and this time they would stand in quarz.
in early March I will be able to publish the results....:redface:
thanks again to those who participated and who would want!

I have one question:
on nx the error type "body type non manifold" has been solved or not?
if you = maybe a day (maybe, maybe) will succeed on solid edge
If not =...

 

[Tequila]

I have one question:
on nx the error type "body type non manifold" has been solved or not?
if you = maybe a day (maybe, maybe) will succeed on solid edge
If not =...

I think it's gonna be a utopia.

 

[stef_design]

I have one question:
on nx the error type "body type non manifold" has been solved or not?
if you = maybe a day (maybe, maybe) will succeed on solid edge
If not =...

I mean, what does that mean?

 

[cacciatorino]

I mean, what does that mean?

in part environment, solidedge (and maybe even nx?) is not able to handle tangent bodies (e.g. ball bearing or roller) so if you try to import a step so made into a part model, the solid is not created but remains a cluster of superifices.

 

[Tequila]

I mean, what does that mean?

or in addition to the bearing mentioned by hunting, if for example you make a hole with finite extension with bottom corner (a classic hole with the tip) and the total depth of the hole coincides with that of the thickness of the extrusion where you apply it, it gives you error and does not do it, you are forced to make the hole shorter even of 1 thousandth because the feature can be created.
If I also have to do a stupid simulation with two parallelepipeds (see attachments) with the tangent edge does not let you do it, we know very well that it is impossible to achieve physically but if I only need a flight simulation of 2 stupid objects I am forced to create a set.

I assume that even nx has this limit, what other cads (swx, inventor) have already solved for years.

 

[ipotemusa]

I mean, what does that mean?

But do you realize?
Who doesn't use sw zimenz manco knows what a mistake like that!

If you draw 2 you look for tangents and try to extrude them with sedge you can't do it.
but if you do a set and 2 cylinders, you will fit them in tangenza no problem.
There are always import problems if you tend to convert a set (example of the bearing) to a part, it does not work.

It is not a question of this.

 

[wired]

But do you realize?
Who doesn't use sw zimenz manco knows what a mistake like that!

If you draw 2 you look for tangents and try to extrude them with sedge you can't do it.
but if you do a set and 2 cylinders, you will fit them in tangenza no problem.
There are always import problems if you tend to convert a set (example of the bearing) to a part, it does not work.

It is not a question of this.

I imagine you use nx to say that siemens software is all the same
any multibody cad manages the situations you described, nx included.
if it's multibody? If yes, then you can do, if not, you have to do the asses. It's not a kiss, it's a cad philosophy and you can't change it with a release, you have to flip the software.

right for clarity, nx does not even differentiate parts from assemblies, the extension is always .prt.
modeling bodies, remains a part, adding components (pointers) , becomes together, and vice versa
super flexible:finger:

 

[Tequila]

I imagine you use nx to say that siemens software is all the same
any multibody cad manages the situations you described, nx included.
if it's multibody? If yes, then you can do, if not, you have to do the asses. It's not a kiss, it's a cad philosophy and you can't change it with a release, you have to flip the software.

right for clarity, nx does not even differentiate parts from assemblies, the extension is always .prt.
modeling bodies, remains a part, adding components (pointers) , becomes together, and vice versa
super flexible:finger:

the underlying problem and for which me and other users of if we beat us is that this cad that uses the same 3d engine of nx (use the same version of parasolid) must have these big differences.
We do not claim that he does the miracles that he does nx, also because otherwise he should retouch to the great rise the price list, but certain shortcomings and problems that we signal for years should settle them.
from what you write about nx to the point it shows that solid edge is the only parametric cad 3d that does not manage the multibody

 

[SolidUser]

I answer everyone.
This type of error is related to parasolid.
all modelers using the parasolid kernel have this limit (solidworks, solid edge, nx, etc...)
who uses acis like catià and inventor do not have this limitation.

apart from this explanation, I don't understand what your problem is.

 

[SolidUser]

in part environment, solidedge (and maybe even nx?) is not able to handle tangent bodies (e.g. ball bearing or roller) so if you try to import a step so made into a part model, the solid is not created but remains a cluster of superifices.

This is just a solid edge problem that doesn't support multibody.
other cads using parasolid, step files with tangences are imported.
However if the bearing was created as a whole, when it matters, you find it as a whole and then you can.
We need to see who exported it.
maybe if it was exported by a cad that uses acis, you find the surfaces because they have a greater precision.

 

[cacciatorino]

all modelers using the parasolid kernel have this limit (solidworks, solid edge, nx, etc...)


apart from this explanation, I don't understand what your problem is.

Sorry I don't understand, solidworks manages the multibody very well....

the problem is that, in the presence of tangent surfaces, if it fails to create the solid and the imported file remains a cluster of surfaces of which the cad does not understand which is the inside or the outside and various other amenities.

 

[SolidUser]

But do you realize?
Who doesn't use sw zimenz manco knows what a mistake like that!

If you draw 2 you look for tangents and try to extrude them with sedge you can't do it.
but if you do a set and 2 cylinders, you will fit them in tangenza no problem.
There are always import problems if you tend to convert a set (example of the bearing) to a part, it does not work.

It is not a question of this.

How do you consider two tangent circles as one object, i.e. merged together?
Am I?
I'm not really.
therefore the logic is correct.
in the axieme works, because they are two distinct objects.
the cad must follow reality, I don't like you to do something impossible.
catià manages to make rays that others do not succeed, but I have strong doubts that they are feasible...who knows what changes.

 

[SolidUser]

Sorry I don't understand, solidworks manages the multibody very well....

the problem is that, in the presence of tangent surfaces, if it fails to create the solid and the imported file remains a cluster of surfaces of which the cad does not understand which is the inside or the outside and various other amenities.

try to create two tangent circles and extrude them.. Always in solidworks.
You'll see that you can't... because I'm inside the same function that tries to keep a fused object.
if instead the two circles are on different sketches, then if you remove the flag of 'join the result', then you can because you enter multibody mode.
but if you try to merge them, then even in this case you can't.
In fact, you can't merge two tangent cylinders.
you can with addition of material (saving).

 

[cacciatorino]

try to create two tangent circles and extrude them.. Always in solidworks.
You'll see that you can't... because I'm inside the same function that tries to keep a fused object.
if instead the two circles are on different sketches, then if you remove the flag of 'join the result', then you can because you enter multibody mode.
but if you try to merge them, then even in this case you can't.
In fact, you can't merge two tangent cylinders.
you can with addition of material (saving).

Yes, that's what I meant: if by default you use the "result unit" and it's not possible to do otherwise. so the classic ball bearing you download from the skf site you have to hold it as (useless) assembly and have 5-6 files in the file system only to be able to mount your bearing in the machine. instead in swx you can manage it as multibody and get all the bearing inside a single sldprt file, which simplifies management.

Now I've had to load a file step of a supplier, and if it treats me as a cluster of superfci so it can't calculate weights, it's in trouble in the table, etc. if it had the multibody I wouldn't have any problem.

 

[Tequila]

I answer everyone.
This type of error is related to parasolid.
all modelers using the parasolid kernel have this limit (solidworks, solid edge, nx, etc...)
who uses acis like catià and inventor do not have this limitation.

apart from this explanation, I don't understand what your problem is.

You have already been answered what the problem is. when importing files from external cads often and gladly you do not need to have all the myriad components, example of the bearing with inner ring + external + balls or rollers + shields + everything you want, but just as long as it is a unique body to have a single file to manage. so if you try to import these files as a single solid body, the cad fails and generates a myriad of surfaces and you cannot manage its weight and for putting them on the table becomes along the work to show these surfaces. if you manage it as a whole, which does not make any sense, then on the main axieme long go of addition of components becomes heavier and slower in loading/saving, manipulating etc.
and all this because of the non-manifold error

 

[SolidUser]

Yes, that's what I meant: if by default you use the "result unit" and it's not possible to do otherwise. so the classic ball bearing you download from the skf site you have to hold it as (useless) assembly and have 5-6 files in the file system only to be able to mount your bearing in the machine. instead in swx you can manage it as multibody and get all the bearing inside a single sldprt file, which simplifies management.

Now I've had to load a file step of a supplier, and if it treats me as a cluster of superfci so it can't calculate weights, it's in trouble in the table, etc. if it had the multibody I wouldn't have any problem.

Have you tried to change the import tolerance?

 

[Er Presidente]

"non manifold" error is not a simple sw error but a philosophical concept.
the "non-constructible" indicates that you are trying to shape a non-real solid.
the cads are divided into two worlds, those who accept the "non manifolds" and those who accept them.
the fact that it is right to be on one side or the other is part of the sphere of personal choices, to consider it a "error" is contrary to logic.
the designer of a cad's "motor" chooses in which half of the world to be, if you don't like to choose another sw that is "on the other side".
designer sw has the right to choose where to stay exactly as you do.

:biggrin:

 

[wired]

....
from what you write about nx to the point it shows that solid edge is the only parametric cad 3d that does not manage the multibody

also proe is not multibody.

 

[wired]

...

the imho discussion is not about combining two tangent bodies, but in being able to manage separate solids in a single file without having to resort to the assembly... .
I think it's logical that 2 tangency bodies cannot be considered one body. .

 

[The_Matrix]

I answer everyone.
This type of error is related to parasolid.
all modelers using the parasolid kernel have this limit (solidworks, solid edge, nx, etc...)
who uses acis like catià and inventor do not have this limitation.

apart from this explanation, I don't understand what your problem is.

catia non uses acis...